US 12,272,245 B2
Bypass assistance
Boaz Mell, Petach Tikva (IL); Adam Harel, Tel Aviv (IL); and Igal Raichelgauz, Tel Aviv (IL)
Assigned to AUTOBRAINS TECHNOLOGIES LTD, Tel Aviv (IL)
Filed by AUTOBRAINS TECHNOLOGIES LTD, Tel Aviv (IL)
Filed on Jul. 23, 2021, as Appl. No. 17/443,318.
Claims priority of provisional application 62/705,974, filed on Jul. 24, 2020.
Prior Publication US 2022/0028273 A1, Jan. 27, 2022
Int. Cl. G08G 1/16 (2006.01); B60W 30/09 (2012.01); B60W 50/14 (2020.01)
CPC G08G 1/166 (2013.01) [B60W 30/09 (2013.01); B60W 50/14 (2013.01)] 17 Claims
OG exemplary drawing
 
1. A method for bypass alert, the method comprises:
obtaining sensed information about an environment of a vehicle, by one or more vehicle sensors associated with the vehicle; wherein the environment of the vehicle comprises a passing lane, a lane separator, and a current lane in which the vehicle is positioned; wherein a driving direction of the passing lane is in an opposite direction to a driving direction of the current lane; wherein the lane separator separates the passing lane from the current lane;
determining whether a driver of the vehicle initiates a bypass of another vehicle, wherein the determining is made by a first machine learning process, and is based on detecting a movement of the vehicle in relation to the lane separator, wherein the driver of the vehicle is a human driver;
determining that the vehicle entered the passing lane;
determining a situation based on a signature of the sensed information;
determining, by a second machine learning process and in response to the situation, a time to collision for the vehicle with respect to an incoming vehicle identified driving at the passing lane towards the vehicle;
determining, based on, at least in part, the determined time to collision and further in accordance with an allowable time to collision, a safety parameter for the vehicle with respect to a completion of the bypass; and
responding to the determined time to collision by switching from a human driver control over the vehicle to an autonomous driving of the vehicle based on the determined safety parameter;
wherein the signature is generated by an iterative signature generation process that comprises first iterations and second iterations, the first iterations are executed using a third neural network and the second iterations are executed without using the third neural network.